For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Mojo package

conv

Convolution kernels (1D, 2D, 3D, transposed).

Shared utilities live here (conv_utils), with platform-specific implementations organized by vendor and architecture:

conv/ - shared utilities and main conv implementations conv/conv.mojo - CPU direct conv, GPU conv via cuDNN/MIOpen conv/conv_transpose.mojo - transposed (deconvolution) for CPU and GPU conv/conv_utils.mojo - ConvShape, ConvPartition, epilogue helpers conv/gpu/nvidia/sm100/ - SM100 (Blackwell) structured Conv2D kernels

Packages

• ​gpu: GPU convolution kernels. Vendor-specific implementations live under nvidia/.

Modules

• ​conv:
• ​conv_transpose:
• ​conv_utils:

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit